<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Marella AI research and guides</title><description>Evidence-first guidance for evaluating private enterprise AI, cited answers, deployment and regulated document workflows.</description><link>https://marella.ai/</link><language>en-gb</language><item><title>How accurate is AI on legal questions?</title><link>https://marella.ai/blog/posts/how-accurate-is-ai-on-legal-questions/</link><guid isPermaLink="true">https://marella.ai/blog/posts/how-accurate-is-ai-on-legal-questions/</guid><description>An accuracy figure means little without the corpus and the questions behind it. What to measure instead, and why the no-answer cases tell you most.</description><pubDate>Tue, 11 Aug 2026 00:00:00 GMT</pubDate></item><item><title>How AI reads a legal document, stage by stage</title><link>https://marella.ai/blog/posts/how-ai-reads-a-legal-document/</link><guid isPermaLink="true">https://marella.ai/blog/posts/how-ai-reads-a-legal-document/</guid><description>The model is the fifth stage of five, and rarely the one that decides whether an answer is any good. What happens before it, and where each step falls over.</description><pubDate>Tue, 11 Aug 2026 00:00:00 GMT</pubDate></item><item><title>Privilege, barriers and AI over your documents</title><link>https://marella.ai/blog/posts/privilege-information-barriers-and-ai/</link><guid isPermaLink="true">https://marella.ai/blog/posts/privilege-information-barriers-and-ai/</guid><description>Vendors answer the information barrier question in one sentence about permissions. That sentence covers three architectures that do not protect you equally.</description><pubDate>Tue, 11 Aug 2026 00:00:00 GMT</pubDate></item><item><title>What in-house legal teams need from AI</title><link>https://marella.ai/blog/posts/what-in-house-legal-teams-need-from-ai/</link><guid isPermaLink="true">https://marella.ai/blog/posts/what-in-house-legal-teams-need-from-ai/</guid><description>The in-house bottleneck is usually finding what you already agreed, not drafting faster. Most of the market is built to sell against the other problem.</description><pubDate>Tue, 11 Aug 2026 00:00:00 GMT</pubDate></item><item><title>What ISO 27001 tells you about an AI vendor</title><link>https://marella.ai/blog/posts/what-iso-27001-tells-you-about-an-ai-vendor/</link><guid isPermaLink="true">https://marella.ai/blog/posts/what-iso-27001-tells-you-about-an-ai-vendor/</guid><description>ISO 27001 is a security claim you can verify before signing. What the certificate covers, what it leaves out, and which questions to ask an AI vendor.</description><pubDate>Wed, 01 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Private RAG, explained for software buyers</title><link>https://marella.ai/blog/posts/private-rag-explained/</link><guid isPermaLink="true">https://marella.ai/blog/posts/private-rag-explained/</guid><description>Private RAG is not a standard assurance label. Which boundaries to check on retrieval, on hosting and on your own data before you sign anything.</description><pubDate>Wed, 17 Jun 2026 00:00:00 GMT</pubDate></item><item><title>Why cited answers beat confident answers</title><link>https://marella.ai/blog/posts/cited-answers-beat-confident-answers/</link><guid isPermaLink="true">https://marella.ai/blog/posts/cited-answers-beat-confident-answers/</guid><description>A citation makes an AI answer checkable, but its presence does not prove the answer is right. How to test what one supports before you rely on it.</description><pubDate>Thu, 28 May 2026 00:00:00 GMT</pubDate></item></channel></rss>